Analytics as Code with Juttle
summary
Argues that operational analytics must be treated as code to get versioning, reviews and reuse; shows concrete problems: Kibana-backed setups balloon to over a hundred charts and require manually updating 117 charts when a statistic changes, lack access control and time‑series math. Presents Juttle as a battle‑tested option (in production over a year, company defunct) that reads from elasticsearch sql and TSDBs, joins sources, alerts to Slack/PagerDuty and renders charts. Notes DataDog as a practical SaaS that exports dashboards to readable JSON for git and review.
AI-written summary · about
chapters (9)
- 0:00Introduction to operational analytics
- 2:26Problems with UI-based analytics
- 7:00Operational analytics at Quidd
- 11:33Limitations of Kibana and Graphana
- 13:38Analytics as code benefits
- 15:23Juttle as a unified solution
- 17:31Data Dog as an alternative
- 19:57Future of analytics tools
- 20:20Q&A on other tools
related talks
33mAccidental DataOps: How we matured our organization with data, and how you can too!
12mHow to integrate analytics into DevOps processes using Git
32mEssential Software Observability for Every Role: Look No Further
27mDashboards Are Dead: How To Democratise Data With Live Analytics
21mJourney from Broken Dashboards to O11y Maturity
24mData Pipelines to Dashboards
38mUnleashing the Power of Analytics Engineering
45mA Practical Guide to Dashboarding